A patent can be thought of as a deal struck between an inventor and a state. The state is willing to grant an inventor a time-limited commercial monopoly for their invention (via a granted patent) in exchange for the patent disclosing the invention and demonstrating that it works. The monopoly is the inventor’s reward for adding to the sum of human knowledge and enhancing scientific progress.
At least that’s the idea. But what happens if the data that supposedly demonstrates the invention cannot be reproduced? Has the inventor then actually fulfilled their side of the bargain? This isn’t a hypothetical situation - as reported below more than 1500 scientists were surveyed and 70% admitted to having tried and failed to reproduce someone else’s experiments. More than 50% tried and failed to reproduce their own experiments. This is a serious issue for science and society.
It is reasonable to assume that the same difficulties with reproduction apply to the data in patent applications. Is this a problem for patentees?
Patent Offices worldwide assess patent applications to check that the disclosures of the inventions therein warrant the grant of patents. For instance, the European Patent Office (EPO) requires that a patent application discloses the invention “in a manner sufficiently clear and complete for it to be carried out by a person skilled in the art”. Lack of sufficient disclosure of the invention is a ground for refusal of patent application and for revocation of a granted patent.
If the data in a patent application cannot be reproduced, then does it follow that the application fails to disclose the invention sufficiently for the skilled person to carry out the invention? And does this mean that a large number of patents are potentially invalid?
Data is not always essential to pass the EPO’s sufficiency of disclosure test, but it is almost always required for inventions relating to a medical use of a product. The EPO typically requires the application to contain data reflecting the claimed therapeutic effect. Clinical trial results or in vivo data does not necessarily have to be present, but there must be some data that reflects the claimed effect.
Opponents can certainly try to attack a patent or patent application on the basis that the data therein cannot be reproduced. Luckily for patentees, the EPO gives them a hefty benefit of the doubt. Firstly, the burden of proof initially lies with an opponent to demonstrate the lack of sufficient disclosure. If an opponent wishes to submit that their attempts to reproduce the data in the patent’s Examples have failed, then the precise methods and conditions of the Examples must be applied. Even then, if the EPO cannot decide which evidence is the most convincing, then they will find in favour of the patentee because the opponent bears the burden of proof but has been unable to prove their point convincingly.
Patentees should not be blasé about this issue however. A similar pro-patentee position should not be expected in the National Courts, where cross-examination and an assessment of witness reliability will also take place.
More than 70% of researchers have tried and failed to reproduce another scientist's experiments, and more than half have failed to reproduce their own experiments. Those are some of the telling figures that emerged from Nature's survey of 1,576 researchers who took a brief online questionnaire on reproducibility in research.